---
license: apache-2.0
dataset_info:
  features:
  - name: xml_sha1
    dtype: string
  - name: declaration_xml
    dtype: string
  - name: declaration_json
    dtype: string
  - name: extracted_text
    dtype: string
  splits:
  - name: train
    num_bytes: 193695685
    num_examples: 10944
  download_size: 35700267
  dataset_size: 193695685
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

We removed all B64 encoded values as they clutter the context without providing meaningful information.